I found it a bit concerning the amount of people missing the forest for trees regarding AI on this site
Why are people arguing sentience and not the fact that this is another attempt to completely remove a sector of the workforce the way Covid was the conduit to remove a complete sector of the outside world? This shit is valued at nearly 200 billion dollars
Because too many people, including self-described leftists, believe in the marketing bullshit that called it "AI" to begin with, which has been a staggering triumph... of marketing bullshit.
I don't want to wander into the weeds with the kind of people that have a "just one more burnt forest bro, just one more dried up lake bro, a sufficiently large dataset will make that one le epic scene in Blade Runner 2049 come true" but it's hard not to when they tend to at the same time with smug misanthropic put-downs about us emotional fearful Luddite meat computers and all that.
Yes, the important part is that this latest means of production is so far a highly wasteful solution in search of a problem... though it did find one problem of the bourgeoisie that it can potentially solve: removing workers entirely from the equation wherever it's crammed in, no matter how sloppy, unreliable, or even hazardous the results.
Contempt for workers, and people in general, is the point. And it's fucking exhausting when the old horse of "what if the treat printer wakes up and is a cyberspace waifu that loves me unconditionally" gets trotted out... which is intentional marketing bullshit by the hype pushers.
It comes up sometimes. But ultimately those who are concerned with AI sentience and rights only care if they look humanoid. If a roomba went on strike you’d punt that shit across the room, or just factory reset it
If a roomba went on strike you’d punt that shit across the room, or just factory reset it
Exocomps, at least as presented and expressed in the fiction, are comrades and should they ever become real, they would deserve rights and recognition.
LLM treat printers are not and never will be fucking Exocomps, however, no matter how big and wasteful their databases might become or how convincingly they can be prompted to say "I LOVE YOU SENPAI" to credulous techbros.
I wish it stuck there but not always. It isn't just lemmyverse bazingas, either. It's kind of saddening that a few older accounts here sometimes do the same. Not naming names; if they come, I'll point at them.
I think you may have the wrong hexbear. Most discussions about llms here are how much they suck and or the awful problems they're going to cause when"ai" code turns out to have non-human readable bugs that no one can figure out.
I've had recent unsolicited and undesired "the treat printers are already approaching human sapience and are only getting better, just build a larger database bro why are you emotional and afraid bro" exchanges with not-naming-names here recently. They may even show up in this thread soon.
I'm right here bro, you can't tell me my LLM isn't sentient when the evidence is clearly there
(Although it's an interesting tangent that we're actually talking about machine sapience - the experiential and emotional dimensions of sensation is just taken for granted as something that comes with human intelligence, even though our wetware is similar to all the other animals)
dude when AI art entered public awareness, there were exhausting struggle sessions on this for months, people saying over and over again "but I suffer from anphatasia! it helps me!", no matter how many times you told them these models are trained on stolen work. I got called very nasty things on here no matter how detailed my replies were. Fuck, It got so bad that one of the largest private trackers for CG content literally banned sharing AI tuts and softs because people shitted up the listings lmao. It was fine when LLMs were relatively niche back in 2020/2021, but the AI boom made it insufferable.
Yea, the AI backlash is a very recent phenomenon here. This site is not beating the tech bro, tech gal, and tech enby allegations.
Another favorite pro AI retort was materialism cannot oppose technological progress as if boiling every lake on the planet for fancy auto complete was a requirement for technological progress.
I quit discussing it when i got called an ableist and/or a luddite for saying AI art is derivative bullshit and theft. Those people will recognise themselves, and yes, I sill despise you. I dont have the appetite for struggle sessions so I usually disengage immediately.
called an ableist and/or a luddite for saying AI art is derivative bullshit and theft.
It’s really funny when people get upset when you don’t consider them an artist when using AI generated stuff. They’ll get indignant and scoff at requiring art to have “soul” and human effort. But why is that so bad? It’s okay. You don’t need to be an artist. You don’t have the skills or drive for it, and that’s fine. Most of us won’t ever be this or that. Generate an AI painting and have fun, and let the artists go do their work.
i use AI trained on my own stuff, and then i heavily remix it, it has a use case but that requires producing actual stuff and knowing what to do with it, something tech bros struggle with.
if you're rich, to just want less humans on earth because it means less chance of a supervirus pandemic murdering them. it's almost their only vulnerability to immortality.
Machines break down, machines clog up, machines can compound mechanical errors. The would-be god-kings may just choke to death on their own bazinga if they hid away and tried Ted Faro style bunker lifestyles.
The most salient argument against "AI" is that we just don't fucking need it. For anything. There is no reason it should exist. The ends 1,000,000% do NOT justify the means, and they never, ever will. That's all I have to say about this shit, and anyone who disagrees with this very basic premise is naive at best and should get out of the way. There are too many here like that, yes. You see assholes posting AI generated meme art here for the gags all the time and I want to scream. Unprincipled morons.
All of these discussions lack specificity. "AI" isn't a monolith and means a ton of different things. Image classification and targeting uses ML and it is very helpful and is used widely. Some TTS and STT applications could be considered AI as they use ML models for continuous improvement. Translation services use ML. Exoplanet hunting uses ML models. Your email filtering uses ML. AI stuff is used everywhere in cybersecurity. "AI=LLM" is missing a ton of nuance and information.